Deep Learning Scientist Intern
Company Details
We develop pioneering technology to empower analysts to take smart actions for a safer world. Our artificial intelligence solutions are based on an automated analysis of geospatial intelligence (GEOINT) workflows. Powering state-of-the-art Machine Learning and Computer Vision, we aggregate information from multi-intelligence sources (Multi-INT), such as imagery (IMINT), electromagnetic (ELINT), or open source (OSINT), to provide our Defence and Intelligence customers with strategic knowledge regarding mission critical areas of interest ensuring a more insightful operating picture.
JOB TARGET
You will join the AI Product team that works on object detection algorithms on images, for Preligens site surveillance product, as a deep learning scientist intern.
The principal team mission is to provide and guarantee operational object detection and classification performances for our site monitoring product.
To that end, you will contribute to one or several projects that aim to develop or improve a specific algorithm. You will collaborate closely with our AI framework team to transform innovative and smart ideas into tools usable by everyone. You will have the opportunity to work with both our internal business experts to understand the needs of our clients, and our data management team to access our vast database of labeled satellite images.
MAIN AREAS OF ACCOUNTABILITY
Contribute to the technical part of the development of an algorithm: choice of the training and testing dataset, application of a new training technique, as well as the integration of the algorithm on the product and its customer delivery.
MAIN ACTIVITIES
- Developing, training and testing the deep learning algorithms using our AI framework
- Trying innovations of our AI R&D team to push forward the performance of our algorithms
- Coming up with new creative and innovative ideas to push forward the performance of our algorithms
- Contributing to the specifications of our AI framework by providing your needs
INFLUENCE ON RESULTS
Must-haves:
- Education in the field of data science / deep learning / computer vision
- At least one practical experience in computer vision / deep learning (can include internships, online courses or personal side projects)
- Knowledge on of DL techniques, especially CNNs for Segmentation and Object Detection
- Good skills in Python
- Experience of building and training DL models in a framework such as Keras, TensorFlow or PyTorch
- Comfortable in a UNIX/Linux environment
- Good communication and team work skills
- Rigorous, creative and meticulous mind
- Willingness to take on challenges, show resiliency and like to always learn new skills
Nice-to-haves:
- Knowledge of some classical Computer Vision techniques
- Proficiency with docker, github
- Experience working in a startup environment
- Experience in people management
- Experience in packaging and delivering models into production
